LVD SCSI - April 18th meeting Comments

Paul D. Aloisi 603-429-8687 aloisi@unitrode.com ALOISI at uicc.com
Mon Apr 7 04:54:03 PDT 1997


* From the SCSI Reflector (scsi at symbios.com), posted by:
* "Paul D. Aloisi 603-429-8687 aloisi at unitrode.com" <ALOISI at uicc.com>
*

There has been a lot of discussion in the last meeting and over
the reflector about Ultra3, and beyond. There are some
speed bumps in the road, but we knew this from the beginning.

It is interesting that most of the discussion is coming from
people who have not been involved in the development meetings
|from the beginning. Skip, keep in mind that talking about Fast-300+
speeds is like talking to the group in 84-85 that were in the 
process of developing the Async SCSI protocol for speed less
than 5 MBs (Mega Bytes per Second) about Fast-20 wide (Ultra Wide)
transfering at 40 MBs. Tell the same person who just got a
personal computer 286 4.7 MHz that he would be able to buy a 200
MHz pentium Pro with a 4 gig for less money than he paid for 286 
with a 10 meg disk drive.

We aren't sure how we will get there, but there are at least 6
major improvements we know of now that will get us well past the
Ultra2 speeds. Keep in mind that we have increased speeds in the past
when the market was ready. The 2 and 4 port Ultra2 SCSI cards
have more bandwidth than the PCI bus. Yes, a Silicon Graphics
machine could use a higher speed SCSI bus, but 99% of your
systems are not ready for IO buses with higher speed unless you
build the adapter on the processor bus. Ultra just came into the market
the end of last year for the masses, Ultra3 isn't needed yet.

Ultra3 should be talked about as a future step, but the April
18th meeting is political by a company that is late to the game
and has not put the effort in from the beginning. I know fiber
channel is more exciting to work on, but from a history perspective
it is a revolutionary change, not an evolutionary change. Revolutionary
steps take 5 years development, 5 years pilot level, then wide 
acceptance in the market place. 15 years after it is started in 
becomes the accepted widely depolied standard. Fiber channel is just
entering the 5 year pilot level. If it is accepted by the industry
we will know in 5 years. Evolutionary steps take 3-4 years to be 
accepted in the market place.  

The April 18th meeting should be an interesting political meeting,
sorry to say it will be the first LVD SCSI meeting that I will not
be attending for personal (Family) reasons. There are at least 6 
different ways that should be explored, not just the one being
purposed to work at Fast-80 and beyond. The proposal is only a minor
benifet to Fast-80, other changes will be much more effective
for higher speeds. Changes that will not effect anything that is 
already designed.

Sorry for the long message,
Paul Aloisi
Unitrode
*
* For SCSI Reflector information, send a message with
* 'info scsi' (no quotes) in the message body to majordomo at symbios.com




More information about the T10 mailing list